This R21 proposal, that is responsive to PA-03-003 "Exploratory Studies in Cancer Detection, Diagnosis and Prognosis," builds on our recent demonstration of the presence of two major genetic haplotypes in the far upstream enhancer region of the prostate specific antigen (PSA) gene that account for 97% of all PSA gene alleles. We also demonstrated that these haplotypes are associated with 30 to 50% differences in serum PSA levels in men without prostatic disease. Our data demonstrate a role for genetic variation on PSA levels in men without prostatic disease. In addition to the use of PSA as a serum marker, several in vitro and experimental model systems have focused on the direct role of PSA as a regulator of prostatic growth factors due to PSA's proteolytic activity. Both positive and negative effects of PSA on prostate cancer growth, differentiation, and aggressiveness have been demonstrated. The experimental systems that show potential regulatory properties of PSA are intriguing but remain untested in humans. Our discovery of PSA promoter SNPs that are associated with increased PSA expression offers a unique opportunity to correlate PSA expression with prostate cancer aggressiveness in men. We hypothesize that genetic differences in PSA expression are associated with differences in prostate cancer aggressiveness. This effect of PSA should be reflected in the histological grade of the tumor and in the long-term follow-up of men treated for prostate cancer. In the proposed study we will associate PSA haplotype with histologic and clinical features of prostate cancer aggressiveness in a well-characterized study population of 1,170 men who had prostatectomies at Stanford University between 1983 and 1998, 990 with a minimum 5-year follow-up. In aim 1A we will correlate PSA promoter genotype with percent tumor volume of Gleason grade 4t5, tumor location, PSA at diagnosis, and total tumor volume. In Aim 1B we will correlate PSA promoter genotype with the risk of biochemical (PSA) failure. If our hypothesis is correct and PSA expression is associated with prostate tumor aggressiveness, then the PSA SNPs we have identified may be important genetic markers that can be used for prostate cancer risk. In addition, the identification of PSA as a modulator of prostate cancer aggressiveness may change the way that the PSA test is used for prostate cancer screening.